Summary: | Subpixel rendering: Adjust cliprect with devicePixelFractionFromRenderer() before painting. | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|
Product: | WebKit | Reporter: | zalan <zalan> | ||||||||
Component: | Layout and Rendering | Assignee: | zalan <zalan> | ||||||||
Status: | RESOLVED FIXED | ||||||||||
Severity: | Normal | CC: | commit-queue, esprehn+autocc, glenn, kondapallykalyan, simon.fraser | ||||||||
Priority: | P2 | Keywords: | InRadar, Regression | ||||||||
Version: | 528+ (Nightly build) | ||||||||||
Hardware: | Unspecified | ||||||||||
OS: | Unspecified | ||||||||||
Attachments: |
|
Description
zalan
2014-07-15 16:12:52 PDT
Created attachment 234963 [details]
Patch
Created attachment 234964 [details]
Patch
Comment on attachment 234964 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=234964&action=review > Source/WebCore/ChangeLog:9 > + When the compositing layer has tiled layering, offsetting the cliprect with the subpixel accumulation "has tiled layering" -> "is tiled". > Source/WebCore/ChangeLog:10 > + leaves a line of one device pixel gap on the non-leftmost tiles. It's because leftmost tiles get leaves single device-pixel gap > Source/WebCore/rendering/RenderLayer.cpp:1697 > + // FIXME: Moving the cliprect by subPixelAccumulation requires to get IntRect enclosing removed from clipping/dirtyPaintRect. "requires removal of the enclosingIntRect code from..." > Source/WebCore/rendering/RenderLayer.cpp:3596 > + if (!paintingInfo.subPixelAccumulation.isEmpty()) > + adjustedClipRect.expand(paintingInfo.subPixelAccumulation); I'm not sure I agree with this expanding here. You're fixing a bug specific to tiled layers with a change that affects clipping in all composited layers, and may actually result in incorrect clipping. Created attachment 235030 [details]
Patch
Comment on attachment 235030 [details] Patch Clearing flags on attachment: 235030 Committed r171165: <http://trac.webkit.org/changeset/171165> All reviewed patches have been landed. Closing bug. |